Vulnerabilities (CVE)

Server-Side Request Forgery (SSRF) vulnerability in Erlang/OTP ftp (ftp_internal module) allows FTP bounce attacks and SSRF via an unvalidated PASV response IP address. The ftp_internal:handle_ctrl_result/2 PASV handler (mode=passive, ipfamily=inet, ftp_extension=false) extracts the IP address from the server's 227 response and passes it directly to gen_tcp:connect/4 without validating it against the control connection peer address. The adjacent EPSV handlers correctly call peername(CSock) to derive the IP from the control connection, but the PASV handler does not. A malicious or compromised FTP server can redirect the client's data connection to an arbitrary internal host and port. On read operations (ftp:ls/1,2, ftp:nlist/1,2, ftp:recv/2,3), data from the redirected target is returned to the caller. On write operations (ftp:send/2,3, ftp:append/2,3), file content is sent to the redirected target. This enables SSRF against internal hosts, cloud metadata endpoints, and FTP bounce attacks against third-party hosts. The vulnerable path is the default configuration (mode=passive, ipfamily=inet, ftp_extension=false). RFC 2577 section 3 explicitly recommends validating the PASV response IP against the control connection peer. The ftp application is deprecated and scheduled for removal in OTP-30. This vulnerability is associated with program files lib/inets/src/ftp/ftp_internal.erl (inets 5.10.4 through 6.5, OTP 17.4 through 20.3) and lib/ftp/src/ftp_internal.erl (ftp 1.0 and later, OTP 21.0 and later). This issue affects OTP from OTP 17.4 before OTP 29.0.2, OTP 28.5.0.2 and OTP 27.3.4.13, corresponding to inets from 5.10.4 before 7.0 and ftp from 1.0 before 1.2.6, 1.2.4.1 and 1.2.3.1.Show less

Exposure of Sensitive Information to an Unauthorized Actor vulnerability in Erlang OTP ssh (ssh_sftpd module) allows File Discovery. The SSH_FXP_READLINK handler in ssh_sftpd sends the raw result of file:read_link/2 to the client without calling chroot_filename/2 to strip the backend root prefix. An authenticated SFTP client can create a symlink inside the chroot pointing to /; ssh_sftpd resolves the target to the absolute backend root and stores it on disk. Reading the symlink back via SSH_FXP_READLINK returns that absolute path, for example /data/sftp, instead of the chrooted value /. The information disclosed is the absolute filesystem path of the SFTP root directory and of any symlink targets within it. No file contents, credentials, or access to paths outside the root directory are obtainable through this issue alone. This vulnerability is associated with program files lib/ssh/src/ssh_sftpd.erl. This issue affects OTP from OTP 17.0 before OTP 27.3.4.13, OTP 28.5.0.2, and OTP 29.0.2, corresponding to ssh from 3.0.1 before 5.2.11.8, 5.5.2.1, and 6.0.1. Whether OTP before OTP 17.0, corresponding to ssh before 3.0.1, is affected is unknown.Show less

Allocation of Resources Without Limits or Throttling vulnerability in elixir-grpc grpc allows unauthenticated attackers to exhaust the BEAM's memory and crash the server by streaming a large or slow-trickle unary request body. 'Elixir.GRPC.Server.Adapters.Cowboy.Handler':read_full_body/3 (lib/grpc/server/adapters/cowboy/handler.ex) accumulates every received chunk into a single growing binary with no size cap. Additionally, when the client omits the grpc-timeout header, the per-chunk read timeout resolves to :infinity, allowing a slow-trickle client to keep the connection alive indefinitely while memory grows. A single connection is sufficient to exhaust server memory and crash the node. This issue affects grpc: from 0.3.0-alpha.2 before 1.0.0.Show less

Deserialization of Untrusted Data and Allocation of Resources Without Limits or Throttling vulnerabilities in elixir-grpc grpc allow unauthenticated attackers to crash the BEAM node via atom table exhaustion and, when a decoded term flows into a call site that invokes it, achieve remote code execution on the server. 'Elixir.GRPC.Codec.Erlpack':decode/2 (lib/grpc/codec/erlpack.ex) calls :erlang.binary_to_term/1 on the raw gRPC message body without the :safe option, no size bound, and no type guard. Any unauthenticated peer that sends a request with Content-Type: application/grpc+erlpack can send a crafted payload that mints arbitrary new atoms (which are never garbage-collected, exhausting the bounded atom table and crashing the VM) or that encodes a fun term which, if applied anywhere downstream, executes attacker-controlled code inside the server process. This issue affects grpc: from 0.4.0 before 1.0.0.Show less

Starlette is a lightweight ASGI framework/toolkit. In versions 1.0.1 and below, when dispatching a request, HTTPEndpoint selects the handler by lowercasing the HTTP method and looking it up as an attribute with getattr, without restricting the lookup to a known set of HTTP verbs. When an HTTPEndpoint subclass is registered through Route(...) without an explicit methods= argument, the route does not constrain the method and every method reaches the endpoint. If a non-standard HTTP method whose lowercased name matches an attribute on the endpoint subclass reaches the endpoint, that attribute is invoked as if it were a request handler. An attacker can use this to reach methods that were never meant to be HTTP handlers, such as internal helpers, without the authorization checks applied by the intended public handler. An application (including Starlette-based frameworks like FastAPI) is affected if it registers an HTTPEndpoint subclass via Route(...) without explicitly setting methods=, and that subclass includes extra methods named like non-standard HTTP verbs that take one request argument and return a response. This issue has been fixed in version 1.1.0.Show less

Network-AI is a TypeScript/Node.js multi-agent orchestrator. In versions 5.7.1 and earlier, the MCP SSE server allows unauthenticated cross-origin MCP tool invocation due to an empty default secret. This issue was partially addressed by CVE-2026-46701 in version 5.4.5 by closing the CORS flaw (with Access-Control-Allow-Origin now set only for localhost origins), but the empty-default-secret flaw described in the title remained: the SSE MCP server still defaulted to an empty secret, _isAuthorized() still returned true when the secret was empty, and a non-loopback bind only produced a warning. As a result, the server still ran fully unauthenticated by default. Any non-browser caller (for example, curl, SSRF, or a 0.0.0.0 bind) could invoke all 22 MCP tools (config_set, agent_spawn, blackboard_write, token_*) with no credentials. This issue was fixed in version 5.7.2.Show less

Remark42 is a self-hosted comment engine for blogs, articles, or any other place where readers can add comments. Versions 1.6.0 through 1.15.0 contain a Cross-Site Scripting (XSS) vulnerability exploitable through content-type spoofing. The Remark42 image proxy fetches an arbitrary remote URL and re-serves the response from Remark42's own origin. During the download phase, the proxy determines whether the resource is an image by inspecting only the Content-Type header advertised by the remote server, never examining the actual bytes; during the serving phase, it instead derives the response Content-Type by sniffing those bytes with http.DetectContentType. An attacker can exploit this inconsistency by hosting a URL that advertises Content-Type: image/png while returning an HTML/JavaScript body: the download check accepts it as an image, the serving path sniffs the body and emits Content-Type: text/html, and the browser renders the attacker-controlled HTML/JavaScript as a document within Remark42's origin. Exploitation requires no Remark42 account on the target instance; the attacker only needs to host the malicious upstream URL and deliver the proxy link to a victim by any means, such as email, direct message, or a link on another website. This issue has been fixed in version 1.16.0.Show less

gin-vue-admin is an AI-assisted basic development platform. In version 2.9.1, an authenticated attacker with access to the code-generation feature and MCP management interface can exploit this vulnerability by injecting attacker-controlled Go source code through POST /autoCode/addFunc, and then invoking POST /autoCode/mcpStart to trigger a rebuild and restart of the standalone MCP service. This allows arbitrary operating system commands to be executed on the server with the privileges of the application process. Successful exploitation may lead to remote code execution (RCE), modification of backend source code or runtime logic, deployment of persistent backdoors, access to or manipulation of application data and configuration, and further impact on local resources running under the same service account or privilege context. The risk is highest in deployments that retain the source tree, allow writes to source files, and support local build or startup of standalone MCP components. In environments using binary-only releases, read-only filesystems, or with local build capabilities removed, the exploitability of the full attack chain is significantly reduced. However, once the online code-generation capability and MCP-hosted startup workflow are enabled, the overall security impact may reach high to critical severity. As of time of publication, it is unknown if a patched version is available. As a workaround, enforce strict allowlist validation on path- and identifier-related fields such as `humpPackageName`, `packageName`, `FuncName`, and `Router`, and only permit safe identifier formats.Show less

FileBrowser Quantum is a free, self-hosted, web-based file manager. Versions prior to 1.3.2-stable, 1.4.0-beta and 1.4.1-beta are vulnerable to Path Traversal through the publicPatchHandler in backend/http/public.go which joins user-controlled fromPath and toPath body fields with the trusted d.share.Path BEFORE the downstream sanitizer runs. Because filepath.Join collapses .. segments during the join, the sanitizer in resourcePatchHandler never sees the traversal and the move/copy/rename operates on a path outside the shared directory. The same root-cause pattern was patched for the bulk DELETE endpoint as CVE-2026-44542 (GHSA-fwj3-42wh-8673), but the PATCH handler with the identical pattern was not updated. A public share link with AllowModify=true is sufficient to exploit this. Anyone holding such a link can move, copy, or rename arbitrary files within the share owner's source root. This issue has been fixed in versions 1.3.3-stable and 1.4.2-beta.Show less

ProxySQL is a proxy for MySQL and its forks, as well as PostgreSQL. In versions 3.0.0 through 3.0.8, ProxySQL's GenAI/MCP `run_sql_readonly` tool violates its documented read-only contract for MySQL targets. The tool validates only the full input string with a substring blacklist and first-keyword allowlist, but then executes the entire SQL string on a backend connection created with `CLIENT_MULTI_STATEMENTS`. As a result, a caller can submit a read-only first statement followed by a side-effecting second statement, such as `SELECT 1; RENAME TABLE ...`. The validator accepts the payload because it starts with `SELECT` and because side-effecting MySQL statements such as `RENAME TABLE`, `SET`, `RESET`, `LOCK TABLES`, and `KILL` are not rejected by the blacklist. In a live MCP runtime test, the `/mcp/query` endpoint accepted a `run_sql_readonly` request. The MCP response reported success for the first `SELECT`, and direct backend verification showed that the table had actually been renamed. This violates the endpoint's read-only security contract and lets an MCP caller perform backend writes or administrative SQL, limited by the configured MCP target account's database privileges. Version 3.0.9 contains a fix. Other operator mitigations include: keeping MCP disabled unless required; setting a non-empty `mcp-query_endpoint_auth` token before exposing `/mcp/query`; restricting MCP listener network exposure; configuring MCP backend target credentials as database-level read-only users; and adding temporary MCP query rules to block obvious multi-statement patterns.Show less

ProxySQL is a proxy for MySQL and its forks, as well as PostgreSQL. Versions 2.0.18 through 3.0.8 have a pre-authentication heap memory corruption vulnerability in the MySQL and PostgreSQL protocol first-read paths. A remote unauthenticated client can declare an oversized first packet length, and ProxySQL passes that attacker-controlled length directly to `recv()` while writing into a fixed 32 KB input queue. Version 3.0.9 patches the issue.Show less

ProxySQL is a proxy for MySQL and its forks, as well as PostgreSQL. In versions 2.0.0 through 3.0.8, the ProxySQL MySQL frontend accepts the `PROXY UNKNOWN <addr> <addr> <port> <port>\r\n` PP1 frame as a well-formed PROXY protocol header. The HAProxy PROXY protocol v1 specification says that when the protocol token is `UNKNOWN`, the receiver MUST ignore any address fields that follow it, because the proxy has declared it cannot determine the client identity. ProxySQL parses those address fields anyway via `sscanf` and writes the spoofed source address into the session's `addr.addr` field. From there it flows directly into the query-rule matcher, where the `client_addr` predicate decides routing and ACL. When `mysql-proxy_protocol_networks = '*'` (the default), any TCP peer can send a PP1 frame and choose any source IP claim. With that, any `mysql_query_rules` row pinned to a `client_addr` value is forgeable: the attacker writes the address they want to match into the PP1 line, and ProxySQL routes their query as if it came from that address. In practice this is a routing and ACL bypass. Real deployments use `client_addr` for read-write splitting (internal apps go to the primary, public traffic to read replicas), per-app schema pinning, and query-filter rules (DDL allowed only from admin CIDR, public queries blocked from dangerous patterns). An attacker that can reach the frontend port can forge their way into any of those routes. Version 3.0.9 patches this issue.Show less

TypeBot is a chatbot builder tool. In versions prior to 3.17.2, SSRF validation is implemented by resolving a hostname once and checking whether the resolved IP belongs to a forbidden range allowing for DNS rebinding bypass. The root cause is a time-of-check to time-of-use gap in the SSRF guard. The validator resolves the hostname and approves it, but the later request path performs a fresh resolution and connects to whatever IP the hostname maps to at that moment. The actual outbound request is then performed later using the original hostname, without pinning the validated IP to the network connection. An attacker who can supply a URL to a public bot that performs a server-side HTTP Request block or server-side script fetch can use DNS rebinding to pass the initial validation and still force the server to connect to a private or metadata address during the real request. This enables server-side access to private network services, cloud metadata endpoints, and other internal HTTP targets that the validator was intended to block. The exact downstream impact depends on the reachable internal services. Concrete consequences include metadata disclosure, access to internal admin panels, credential theft from metadata services, and further compromise through internal-only HTTP interfaces. This issue has been fixed in version 3.17.2.Show less

vLLM is an inference and serving engine for large language models (LLMs). From 0.3.0 until 0.22.0, a vulnerability in ASGI web servers and starlette's trust on those web servers enables an authentication bypass of the OpenAI API AuthenticationMiddleware. It allows to use the API without providing the configured VLLM_API_KEY or --api-key. This vulnerability is fixed in 0.22.0.Show less

Traccar Client is a GPS tracking mobile app for sending location updates to private servers using the open-source Traccar platform. In versions 9.7.19 and below, a single crafted deep link can silently hijack all GPS tracking parameters and redirect telemetry to an attacker-controlled server. The app registers a custom org.traccar.client://config deep-link scheme that silently writes attacker-supplied parameters (server URL, device ID, accuracy, distance, and interval) into the app's persistent configuration with no confirmation, notification, or visual indication. A single crafted link delivered via SMS, email, a webpage, or any installed app can therefore reconfigure the app the moment the victim taps it, with no special permissions required. As a result, an attacker can covertly redirect all of the victim's GPS telemetry to their own server at maximum precision and frequency, and the change persists across restarts. This gives the attacker continuous, real-time tracking of the victim's location. This issue has been fixed in version 9.7.20.Show less

nanobot is a personal AI assistant. In versions 0.1.5.post3 and prior, the WhatsApp bridge in bridge/src/whatsapp.ts constructs a filesystem path using the fileName field from an incoming WhatsApp document message without sanitization. The WhatsApp bridge downloads media attachments and writes them to disk using a filename derived from the sender's message via documentMessage.fileName, which is concatenated with a prefix and its raw value is passed directly to path.join(mediaDir, outFilename). Node.js path.join resolves .. components, allowing an attacker to escape the intended media/ directory by sending a document with a crafted fileName such as ../../../.ssh/authorized_keys. Because the attacker also controls the file content (the downloaded buffer), this is a write-anywhere primitive — both path and content are attacker-controlled. A fix for this issue is planned for version 0.1.5.post4.Show less

radvd is a router advertisement daemon for IPv6. Prior to version 2.21, the `radvdump` utility shipped with radvd contains a stack buffer overflow in the Route Information option parser. When processing a crafted ICMPv6 Router Advertisement, `print_ff()` copies up to 2032 bytes from attacker-controlled packet data into a 16-byte `struct in6_addr` on the stack, overflowing by up to 2016 bytes. Note that the main `radvd` daemon is not affected by the vulnerability. Version 2.21 patches the issue.Show less

i18next-http-middleware is a middleware to be used with Node.js web frameworks like express or Fastify and also for Deno. In versions prior to 3.9.7, the missingKeyHandler blocked the literal request-body keys __proto__, constructor, and prototype (added in 3.9.3, see GHSA-5fgg-jcpf-8jjw), but did not reject dotted variants such as "__proto__.polluted". Downstream backends that split the missing-key string on a configured keySeparator (notably i18next-fs-backend ≤ 2.6.5) hand these keys to an unguarded setPath() walker that writes to Object.prototype. Applications that expose missingKeyHandler to untrusted input AND use i18next-fs-backend ≤ 2.6.5 are directly exploitable for remote prototype pollution. Other downstream backends that split the missing-key string the same way may be similarly affected. Depending on the host application, polluted prototype properties may cause crashes, corrupted translation behaviour, configuration poisoning, or bypasses of property-based security checks. This issue has been fixed in version 3.9.7. If developers cannot upgrade immediately, they should do the following: do not expose missingKeyHandler to untrusted users (mount it behind authentication, or remove the route), add a request-body filter ahead of the handler that rejects any top-level key containing __proto__, constructor, or prototype after splitting on their configured keySeparator, and disable missing-key persistence (saveMissing: false) when accepting writes from untrusted input.Show less